[Pw_forum] Structural relaxation in electric field

[Joydeep Bhattacharjee]

Hi

Thanks Stefano for your reply regarding the charged
cluster relaxation.

 I now have a problem with structural relaxation
of a cluster in the presence of electric fields as well.
For electric field in one direction the rlaxations takes place
nicely, but if I reverse the field the relaxation "converges"
prematurely when nothing has actually converged, except
the "new trust radius". Below is the relevant excerpt from the
output file:

- - - - - -
- - - - - -

!    total energy              =  -333.72126191 ryd
     estimated scf accuracy    <     0.00000001 ryd

     band energy sum           =   -76.44745596 ryd
     one-electron contribution =  -907.39503374 ryd
     hartree contribution      =   445.49262839 ryd
     xc contribution           =   -46.19803489 ryd
     ewald contribution        =   171.58360056 ryd
     electric field correction =     2.79557778 ryd
     correction for metals     =     0.00000000 ryd

     convergence has been achieved

     Forces acting on atoms (Ry/au):

     atom   1 type  1   force =     0.00000000    0.00000000    0.00000000
     atom   2 type  1   force =     0.00000000    0.00000000    0.00000000
     atom   3 type  1   force =     0.00000000    0.00000000    0.00000000
     atom   4 type  2   force =     0.00000000    0.00000000    0.00000000
     atom   5 type  2   force =     0.00000000    0.00000000    0.00000000
     atom   6 type  3   force =    -0.23800154    0.00299140    0.00514244
     atom   7 type  3   force =    -0.23517359   -0.00315217   -0.00557059

     Total force =     0.334705     Total SCF correction =     0.000274

     number of scf cycles    =   3
     number of bfgs steps    =   1

     energy old              =    -343.2789722890 ryd
     energy new              =    -333.7212619063 ryd

     CASE: energy_new > energy_old

     new trust radius        =       0.0000000000 bohr

     bfgs converged in   3 scf cycles and   1 bfgs steps

     Final energy            =    -343.2789722890 ryd

     Saving the approximate inverse hessian

CELL_PARAMETERS (alat)
   1.000000000   0.000000000   0.000000000
   0.000000000   1.000000000   0.000000000
   0.000000000   0.000000000   1.000000000
ATOMIC_POSITIONS (bohr)
- - - - - - -
- - - - - - -

     Writing file WS3.pun       for program phonon
- - - - - - -
- - - - - - -
*********************************************************************

The input file for a test efield:


 &control
    calculation='relax'
    restart_mode='from_scratch',
    prefix='WS3',
    pseudo_dir = '/export/joydeep/PWSCF/pseudi',
    outdir='/export/joydeep/PWSCF/Mo2S3_H2onMo_efield/tmp',
    tefield= .TRUE.
 /
 &system
    ibrav=0, nat=7, ntyp=3,
    occupations='smearing', smearing='methfessel-paxton', degauss=0.004,
    ecutwfc = 25.0, ecutrho = 130.0,
    edir = 1, emaxpos = 0.6, eopreg = 0.05, eamp = 0.02
 /
 &electrons
    conv_thr =  1.0d-7
    mixing_beta = 0.4
 /
 &ions
    upscale=10
 /
ATOMIC_SPECIES
 S 6.0 S.pbe-van_bm.UPF
 Mo 14.0 Mo.pbe-van_uvw.UPF
 H 1.0 H.pbe-van_bm.UPF
ATOMIC_POSITIONS {bohr}
- - - -
CELL_PARAMETERS {cubic}
 30.0  0.0  0.0
  0.0 30.0  0.0
  0.0  0.0 30.0
K_POINTS {Gamma}



**********************************************************************

Thanks in avance for your advise.

Joydeep
-------------- next part --------------
An HTML attachment was scrubbed...
URL: /pipermail/attachments/20061115/322d3464/attachment.htm